Metabolic bone disease, particularly osteoporosis, is a serious challenge to our health system. In the United States, treatment costs for osteoporosis is currently estimated at $10-$18 billion and expected to double over the next decade as the population ages. Currently, fracture risk is diagnosed by bone density measurements. Substantial improvement in the determination of fracture risk can be gained through knowledge of both bone density and trabecular bone (TB) structural integrity. The fabric tensor provides both a measure of bone density (scalar) and the structural orientation of TB (2nd-rank tensor). Magnetic resonance imaging provides a non-invasive means of examining both bone density and trabecular bone microstructure. Yet, technical limitations of in vivo MRI, the achievable signal-to-noise (SNR) and field-of view (FOV) within a reasonable scan time prevent an accurate determination of the fabric tensor. We propose that the TB fabric tensor and its structural anisotropy can be quantified in vivo, thereby improving the diagnostic sensitivity to disease progression and treatment efficacy. Our long-term specific aims include (i) implementation of parallel imaging at higher field strengths for the acquisition of a larger TB volume for improved analysis of the fabric tensor without SNR degradation, (ii) development of novel image processing techniques robust to noise and partial-voluming for the characterization of TB structure, and (iii) validation of the fabric tensor through comparison to finite-element calculations of theTB stiffness tensor from high resolution datasets.